The stereochemical course of group II intron self-splicing

Phosphoester Hydrolysis and Intramolecular Transesterification of Ribonucleoside 2'- and 3'-Phosphoromonothioate Triesters: Kinetics and Mechanisms for the Reactions of 5'-O-Methyluridine 2'- and 3'-Dimethylphosphoromonothioates

The hydrolytic reactions of the monothioate analogs of 5'-O-methyluridine 2'- and 3'-dimethylphosphates have been followed over a wide acidity range, H(0) = -1.7 ([HCl] = 5 mol L(-)(1)) to pH 9. Two reactions were found to compete: mutual interconversion of the 2'- and 3'-isomers and phosphoester hydrolysis to a mixture of phosphorothioate diesters, viz., the R(P) and S(P) diastereomers of 2',3'-cyclic thiophosphate and 2'/3'-monomethylthiophosphates (i.e., three pairs of diastereomers). No marked desulfurization could be observed. The interconversion and hydrolysis both show first-order dependence of rate on acidity at pH < 0, the isomerization being 3-4 times as fast as the phosphoester hydrolysis. Under less acidic conditions, the hydrolysis remains pH-independent up to pH 7, while the isomerization becomes hydroxide-ion-catalyzed (first-order in [OH(-)]) already at pH 2. The hydrolysis is susceptible to general base catalysis in carboxylic acid buffers, the Brönsted beta value being 0.8. In contrast, no conclusive evidence for buffer-catalyzed isomerization could be obtained. All these reactions are suggested to proceed via a pentacoordinated thiophosphorane intermediate, obtained at pH < 1 by an attack of the neighboring hydroxy function on a protonated (monocationic) thiophosphate group and at pH > 2 by an attack of a deprotonated hydroxy function (oxyanion) on a neutral thiophosphate. The monocationic intermediate (pH < 1) may collapse to hydrolysis and isomerization products without further catalysis (departure of alcohol). The monoanionic thiophosphorane (pH > 2) also gives isomerization products without catalysis (departure of 2'/3'-oxyanion), whereas breakdown to the hydrolysis products needs either a specific or a general acid catalysis process (departure of methanol). Accordingly, the observed general-base-catalyzed hydrolysis most likely consists of consecutive specific base/general acid catalysis. The phosphorothioate triesters studied are, under very acidic conditions, more than 2 orders of magnitude more stable than their oxyphosphate counterparts, whereas the rate-retarding "thio effect" (k(P)(=)(O)/k(P)(=)(S)) is much smaller with the hydroxide ion-catalyzed reactions (ca. 4) and almost negligible with the pH-independent hydrolysis.

Trans-activation of group II intron splicing by nuclear U5 snRNA

Similarities between RNA splicing during autocatalytic excision of group II introns and pre-mRNA processing led to the hypothesis that group II introns might be the evolutionary predecessors of spliceosomal small nuclear RNAs. The ID3 subdomain stem-loop structure of group II introns, the proposed analogue of the spliceosomal U5 snRNA, is thought to be essential for 5' splice site recognition and anchoring of the free 5' exon. Using the group II intron bI1 we have analysed the role of ID3 in splicing. In the absence of ID3 the 5' splice site was recognized accurately and efficiently, but exon anchoring was greatly reduced. This step was restored in the presence of RNA fragments consisting of either the terminal stem-loop structure of ID3 or spliceosomal U5 snRNA. This suggests that the predominant role of both RNAs is to anchor the 5' exon during exon ligation. Furthermore, as U5 complements for the loss of ID3, a similar network of structural RNAs may form the catalytic core of both group II introns and spliceosomes.

Mechanistic aspects of enzymatic catalysis: lessons from comparison of RNA and protein enzymes

A classic approach in biology, both organismal and cellular, is to compare morphologies in order to glean structural and functional commonalities. The comparative approach has also proven valuable on a molecular level. For example, phylogenetic comparisons of RNA sequences have led to determination of conserved secondary and even tertiary structures, and comparisons of protein structures have led to classifications of families of protein folds. Here we take this approach in a mechanistic direction, comparing protein and RNA enzymes. The aim of comparing RNA and protein enzymes is to learn about fundamental physical and chemical principles of biological catalysis. The more recently discovered RNA enzymes, or ribozymes, provide a distinct perspective on long-standing questions of biological catalysis. The differences described in this review have taught us about the aspects of RNA and proteins that are distinct, whereas the common features have helped us to understand the aspects that are fundamental to biological catalysis. This has allowed the framework that was put forth by Jencks for protein catalysts over 20 years ago (1) to be extended to RNA enzymes, generalized, and strengthened.

Length changes in the joining segment between domains 5 and 6 of a group II intron inhibit self-splicing and alter 3' splice site selection

Domain 5 (D5) and domain 6 (D6) are adjacent folded hairpin substructures of self-splicing group II introns that appear to interact within the active ribozyme. Here we describe the effects of changing the length of the 3-nucleotide segment joining D5 to D6 [called J(56)3] on the splicing reactions of intron 5 gamma of the COXI gene of yeast mitochondrial DNA. Shortened variants J(56)0 and J(56)1 were defective in vitro for branching, and the second splicing step was performed inefficiently and inaccurately. The lengthened variant J(56)5 had a milder defect-splicing occurred at a reduced rate but with correct branching and a mostly accurate 3' splice junction choice. Yeast mitochondria were transformed with the J(56)5 allele, and the resulting yeast strain was respiration deficient because of ineffective aI5 gamma splicing. Respiration-competent revertants were recovered, and in one type a single joiner nucleotide was deleted while in the other type a nucleotide of D6 was deleted. Although these revertants still showed partial splicing blocks in vivo and in vitro, including a substantial defect in the second step of splicing, both spliced accurately in vivo. These results establish that a 3-nucleotide J(56) is optimal for this intron, especially for the accuracy of 3' splice junction selection, and indicate that D5 and D6 are probably not coaxially stacked.

Ribonuclease P (RNase P) RNA is converted to a Cd(2+)-ribozyme by a single Rp-phosphorothioate modification in the precursor tRNA at the RNase P cleavage site

To study the cleavage mechanism of bacterial Nase P RNA, we have synthesized precursor tRNA substrates carrying a single Rp- or Sp-phosphorothioate modification at the RNase P cleavage site. Both the Sp- and the Rp-diastereomer reduced the rate of processing by Escherichia coli RNase P RNA at least 1000-fold under conditions where the chemical step is rate-limiting. The Rp-modification had no effect and the Sp-modification had a moderate effect on precursor tRNA ground state binding to RNase P RNA. Processing of the Rp-diastereomeric substrate was largely restored in the presence of the "thiophilic" Cd2+ as the only divalent metal ion, demonstrating direct metal ion coordination to the (pro)-Rp substituent at the cleavage site and arguing against a specific role for Mg(2+)-ions at the pro-Sp oxygen. For the Rp-diastereomeric substrate, Hill plot analysis revealed a cooperative dependence upon [Cd2+] of nH = 1.8, consistent with a two-metal ion mechanism. In the presence of the Sp-modification, neither Mn2+ nor Cd2+ was able to restore detectable cleavage at the canonical site. Instead, the ribozyme promotes cleavage at the neighboring unmodified phosphodiester with low efficiency. Dramatic inhibition of the chemical step by both the Rp- and Sp-phosphorothioate modification is unprecedented among known ribozymes and points to unique features of transition state geometry in the RNase P RNA-catalyzed reaction.

Uridine insertion into preedited mRNA by a mitochondrial extract from Leishmania tarentolae: stereochemical evidence for the enzyme cascade model

An RNA editing-like internal uridine (U) incorporation activity (G. C. Frech, N. Bakalara, L Simpson, and A. M. Simpson, EMBO J. 14:178-187, 1995) and a 3'-terminal U addition activity (N. Bakalara, A. M. Simpson, and L. Simpson, J. Biol. Chem. 264:18679-18686, 1989) have been previously described by using a mitochondrial extract from Leishmania tarentolae. Chiral phosphorothioates were used to investigate the stereoconfiguration requirements and the stereochemical course of these nucleotidyl transfer reactions. The extract utilizes (SP)-alpha-S-UTP for both 3' and internal U incorporation into substrate RNA. The internal as well as the 3' incorporation of (SP)-alpha-S-UTP proceeds via inversion of the stereoconfiguration. Furthermore, internal U incorporation does not occur at sites containing thiophosphodiesters of the RP configuration. Our results are compatible with an enzyme cascade model for this in vitro U insertion activity involving sequential endonuclease and uridylyl transferase directly from UTP and RNA ligase steps and are incompatible with models involving the transfer of U residues from the 3' ends of guide RNAs.

Hydrolysis and Desulfurization of the Diastereomeric Phosphoromonothioate Analogs of Uridine 2',3'-Cyclic Monophosphate

Hydrolyses of the two diastereomeric phosphoromonothioate analogs of uridine 2',3'-cyclic monophosphate [(R(P))- and (S(P))-2',3'-cUMPS] at 363.2 K have been followed by HPLC over pH-range 0-12. In aqueous alkali (pH > 9) only base-catalyzed endocyclic phosphoester hydrolysis to a nearly equimolar mixture of uridine 2'- and 3'-phosphoromonothioates (2'- and 3'-UMPS) takes place, analogously to the hydrolysis of uridine 2',3'-cyclic monophosphate (2',3'-cUMP). The (R(P))- and (S(P))-2',3'-cUMPS are hydrolyzed 50 and 30%, respectively, more slowly than 2',3'-cUMP. Under neutral and acidic conditions, desulfurization of the cyclic thiophosphates to 2',3'-cUMP competes with the phophoester hydrolysis, both reactions being acid-catalyzed at pH < 5. The desulfurization is most pronounced in strongly acidic solutions ([HCl] > 0.1 mol L(-)(1)), where more than 90% of the starting material is degraded via this route. At pH < 2, the thioates are considerably, i.e., more than 1 order of magnitude, more stable than 2',3'-cUMP. While the hydrolysis of 2',3'-cUMP is second-order in hydronium-ion concentration, that of 2',3'-cUMPS exhibits a first-order dependence. The reactivities of the two diastereomers are comparable with each other over the entire pH-range studied, the most significant difference being that the pH-independent desulfurization at pH > 5 is with the R(P)-isomer 5-fold faster than with the S(P)-isomer. In contrast to 2',3'-cUMP, depyrimidination of the starting material (i.e., release of the uracil base) competes with the hydrolysis of the thiophosphate moiety under neutral conditions (pH 6-8).

Group II introns: elaborate ribozymes

Group II introns are found in organelle genomes of plants, fungi and algae as well as in some bacteria. Some group II introns have been shown to self-splice in vitro and thus constitute examples of ribozymes. Their splicing pathway is analogous to the splicing pathway of nuclear pre-mRNA introns. They thus constitute simple models to analyze RNA catalysis of this type of splicing reactions. In this review article, I will summarize our current state of understanding of the ribozyme activity of group II introns and show that their large size correlates with their ability to perform complex tasks. After discussing the similarities found between group II and nuclear pre-mRNA introns, I will briefly evoke how the ribozyme activity of group II introns might be involved in their transposition at the DNA level.

Evidence for the late origin of introns in chloroplast genes from an evolutionary analysis of the genus Euglena

The origin of present day introns is a subject of spirited debate. Any intron evolution theory must account for not only nuclear spliceosomal introns but also their antecedents. The evolution of group II introns is fundamental to this debate, since group II introns are the proposed progenitors of nuclear spliceosomal introns and are found in ancient genes from modern organisms. We have studied the evolution of chloroplast introns and twintrons (introns within introns) in the genus Euglena. Our hypothesis is that Euglena chloroplast introns arose late in the evolution of this lineage and that twintrons were formed by the insertion of one or more introns into existing introns. In the present study we find that 22 out of 26 introns surveyed in six different photosynthesis-related genes from the plastid DNA of Euglena gracilis are not present in one or more basally branching Euglena spp. These results are supportive of a late origin for Euglena chloroplast group II introns. The psbT gene in Euglena viridis, a basally branching Euglena species, contains a single intron in the identical position to a psbT twintron from E.gracilis, a derived species. The E.viridis intron, when compared with 99 other Euglena group II introns, is most similar to the external intron of the E.gracilis psbT twintron. Based on these data, the addition of introns to the ancestral psbT intron in the common ancester of E.viridis and E.gracilis gave rise to the psbT twintron in E.gracilis.

Group II intron ribozymes that cleave DNA and RNA linkages with similar efficiency, and lack contacts with substrate 2'-hydroxyl groups

BACKGROUND

Group II introns are self-splicing RNAs that have mechanistic similarity to the spliceosome complex involved in messenger RNA splicing in eukaryotes. These autocatalytic molecules can be reconfigured into highly specific, multiple-turnover ribozymes that cleave oligonucleotides in trans. We set out to use a simplified system of this kind to study the mechanism of cleavage.

RESULTS

Unlike other catalytic RNA molecules, the group II ribozymes cleave DNA linkages almost as readily as RNA linkages. One ribozyme variant cleaves DNA linkages with an efficiency comparable to that of restriction endonuclease EcoRI. Single deoxynucleotide substitutions in the substrate showed that the ribozymes bind substrate without engaging 2'-hydroxyl groups.

CONCLUSIONS

The ribose 2'-hydroxyl group at the cleavage site has little role in transition-state stabilization by group II ribozymes. Substrate 2'-hydroxyl groups are not involved in substrate binding, suggesting that only base-pairing is required for substrate recognition.

Stereochemical selectivity of group II intron splicing, reverse splicing, and hydrolysis reactions

We have previously shown, using phosphorothioate substitutions at splice site, that both transesterification steps of group II intron self-splicing proceed, by stereochemical inversion, with an Sp but not an Rp phosphorothioate. Under alternative reaction conditions or with various intron fragments, group II introns can splice following hydrolysis at the 5' splice site and can also hydrolyze the bond between spliced exons (the spliced-exon reopening reaction). In this study, we have determined the stereochemical specificities of all of the major model hydrolytic reactions carried out by the aI5 gamma intron from Saccharomyces cerevisiae mitochondria. For all substrates containing exon 1 and most of the intron, the stereospecificity of hydrolysis is the same as for the step 1 transesterification reaction. In contrast, the spliced-exon reopening reaction proceeds with an Rp but not an Sp phosphorothioate at the scissile bond, as does true reverse splicing. Thus, by stereochemistry, this reaction appears to be related to the reverse of step 2 of self-splicing. Finally, a substrate RNA that contains the first exon and nine nucleotides of the intron, when reacted with the intron ribozyme, releases the first exon regardless of the configuration of the phosphorothioate at the 5' splice site, suggesting that this substrate can be cleaved by either the step 1 or the step 2 reaction site. Our findings clarify the relationships of these model reactions to the transesterification reactions of the intact self-splicing system and permit new studies to be interpreted more rigorously.

Catalytically critical nucleotide in domain 5 of a group II intron

Domain 5 (D5) is a small hairpin structure within group II introns. A bimolecular assay system depends on binding by D5 to an intron substrate for self-splicing activity. In this study, mutations in D5 identify two among six nearly invariant nucleotides as being critical for 5' splice junction hydrolysis but unimportant for binding. A mutation at another site in D5 blocks binding. Thus, mutations can distinguish two D5 functions: substrate binding and catalysis. The secondary structure of D5 may resemble helix I formed by the U2 and U6 small nuclear RNAs in the eukaryotic spliceosome. Our results support a revision of the previously proposed correspondence between D5 and helix I on the basis of the critical trinucleotide 5'-AGC-3' present in both. We suggest that this trinucleotide plays a similar role in promoting the chemical reactions for both splicing systems.